Zhipeng Jia
Member Of Technical Staff at Anthropic
Seattle, Washington, United States
Summary
👤
Senior🎓
Top SchoolZhipeng Jia is a systems-focused ML engineer with 11 years of experience building high-performance infrastructure for LLMs and TPUs, currently a Member of Technical Staff at Anthropic after leading ML systems research at Google. His work spans disaggregated prefill/decode serving, activation-sparse LLM inference on low-profile hardware, and novel tensor-parallel algorithms to improve distributed matmul efficiency. A PhD from UT Austin with multiple top-tier systems publications (NSDI, ASPLOS, SOSP), he brings research rigor to production challenges and has contributed practical fixes to widely used open-source projects like RocksDB. Based in Seattle, he pairs deep systems and distributed-memory expertise with hands-on backend and database engineering, often optimizing edge cases—compiler warnings and subtle sub-compaction bugs—that improve long-term maintainability.
11 years of coding experience
7 years of employment as a software developer
Doctor of Philosophy - PhD Computer Science, Doctor of Philosophy - PhD Computer Science at The University of Texas at AustinBachelor’s Degree Computer Science, Bachelor’s Degree Computer Science at Tsinghua University
Chinese, English
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (10)
databases10
storage-engine10
c-language10
cprogramming-language10
database10
data-structure9
algorithm9
data-structures9
algorithms9
debugging8
Programming languages (7)
Github contributions (5)
facebook/rocksdb
Dec 2015 - Dec 2015
A library that provides an embeddable, persistent key-value store for fast storage.
Role in this project:
Back-end Developer / Database Engineer
Back-end Developer / Database EngineerContributions:8 commits, 7 PRs, 2 comments in 9 days
Contributions summary:Zhipeng primarily focused on fixing typos, compiler warnings, and improving code efficiency within the RocksDB codebase. Their contributions involved code modifications across multiple files, including corrections in memtable, compaction picker, and database test files. The user also addressed issues related to data structures like sorting vectors instead of sets and fixed computations during the sub-compaction process. Additionally, the user addressed compiler warnings related to unused std::move calls, which improved overall code quality and maintainability.
key-value-storedatabasestorage-engine
ut-osa/nightcore
Feb 2020 - Aug 2021
Nightcore: Efficient and Scalable Serverless Computing for Latency-Sensitive, Interactive Microservices [ASPLOS '21]
Contributions:292 commits, 1 push, 11 comments in 1 year 5 months
microservicesserverless
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.